Autologous cell replacement: a noninvasive AI approach to clinical release testing.

Area of Science:

• Regenerative Medicine
• Stem Cell Biology
• Bioengineering

Background:

• Human induced pluripotent stem cells (iPSCs) offer an ethical alternative to embryonic stem cells for regenerative therapies.
• Increasing use of iPSCs for generating retinal cells (photoreceptors, RPE, choroidal endothelium) necessitates scalable manufacturing.
• Current manual processes for cell manufacturing require simplification and automation for clinical efficacy.

Purpose of the Study:

• To develop a noninvasive method for monitoring the maturation and function of iPSC-derived grafts.
• To utilize artificial intelligence for predicting donor cell identity and evaluating graft quality before transplantation.
• To enable automated, high-throughput validation of iPSC-derived products for cell replacement therapies.

Main Methods:

• Combined quantitative bright-field microscopy with artificial intelligence (deep neural networks and machine learning).
• Developed algorithms to noninvasively assess iPSC-derived graft maturation, predict cell identity, and evaluate function.
• Implemented a system for preemptive identification and removal of abnormal grafts prior to transplantation.

Main Results:

• The AI-powered microscopy approach successfully monitored iPSC-derived graft maturation.
• The method accurately predicted donor cell identity and evaluated graft function.
• Abnormal grafts were effectively identified and excluded, ensuring product quality.
• The developed approach demonstrated transferability, cost-effectiveness, and high throughput.

Conclusions:

• AI-driven microscopy provides a noninvasive, efficient, and scalable solution for iPSC-derived graft assessment.
• This technology is crucial for advancing autologous cell replacement therapies by ensuring graft quality and safety.
• The method serves as a valuable tool for primary product validation in iPSC manufacturing.